A Systematic Account of Machine Moral Agency

Abstract

In this paper, I develop a preliminary framework that permits groups (or ‘systems’) to be moral agents. I show that this has advantages over traditional accounts of moral agency when applied to cases where machines are involved in moral actions. I appeal to two thought experiments to show that the traditional account can lead us to counterintuitive consequences. Then I present what I call the ‘systematic account’ which I argue avoids these counterintuitive consequences. On my account, machines can be partial moral agents currently fulfilling some but not all of the conditions required for moral agency. Thus, when a machine is part of a group of agents, it can be part of a system that is a moral agent. This framework is a useful starting point as it preserves aspects of traditional accounts of moral agency while also including machines in our moral deliberations.